Abstract
Odor identification impairment is a feature of several neurodegenerative disorders. Although neurodegenerative changes in the frontotemporal lobar degeneration (FTLD) subtypes involve areas important for olfactory processing, data on olfactory function in these patients are limited. An 18-item, multiple-choice odor identification test developed at our memory clinic, the Motol Hospital smell test, was administered to 9 patients with behavioral variant frontotemporal dementia, 13 patients with the language variants, primary nonfluent aphasia (n = 7) and semantic dementia (n = 6), and 8 patients with progressive supranuclear palsy. Compared to the control group (n = 15), all FTLD subgroups showed significant impairment of odor identification (P < .05). The differences between the FTLD subgroups were not significant. No correlation between odor identification and neuropsychological tests results was found. Our data suggest that odor identification impairment is a symptom common to FTLD syndromes, and it seems to be based on olfactory structure damage rather than cognitive decline.
Keywords
Introduction
Olfactory deficit is a frequent finding in several neurodegenerative disorders. It has been well documented in Parkinson’s 1 -3 and Huntington’s diseases, 4,5 as well as in Alzheimer’s disease (AD), 6,7 in which the impairment of odor identification is one of the earliest symptoms. 8,9 On the other hand, only limited data are so far available on olfactory functioning in frontotemporal lobar degeneration (FTLD).
Frontotemporal lobar degeneration represents a heterogeneous group of progressive neurodegenerative disorders. 10 -12 Clinically, FTLD is classified into behavioral variant frontotemporal dementia (bvFTD), characterized by prevailing behavioral and dysexecutive symptoms, and the language variants collectively termed primary progressive aphasia (PPA), which is further subdivided into semantic dementia (SD), characterized by the dominant loss of semantic knowledge, and primary nonfluent aphasia (PNFA), characterized by gradual expressive aphasia. 13 Recently, 2 forms of atypical parkinsonian syndrome, progressive supranuclear palsy (PSP) and corticobasal degeneration (CBD), were included into the FTLD family, reflecting their common pathological basis and the presence of cognitive deficit. 14 -16
Brain structures crucial for olfactory processing are localized in the frontal and temporal lobes. 17 -22 Olfactory information passes from the epithelium of the nasal cavity directly to the olfactory bulbs located at the ventral aspect of the frontal lobe. It then moves to the olfactory cortex, involving structures of the mesial temporal area and the basal frontal lobe (piriform cortex, amygdala, uncus, prepiriform area, and entorhinal area), although the neuroanatomy of olfactory cortex is not clearly defined. From the olfactory cortex, olfactory information passes to the orbitofrontal cortex, partially through the mediodorsal thalamic nucleus. The olfactory cortex is also connected to the hippocampal formation, other limbic structures, and neocortical association areas.
Recognition of olfactory stimuli is a multilevel process. Odor detection (measurable as odor threshold) is the ability to smell the odor and is considered to be rather a function of peripheral structures and olfactory bulbs. Odor discrimination (the ability to distinguish different odors) and especially odor identification (tested usually by matching the right name or picture to the odor) rely more on cortical structures of frontal and temporal lobes. 22,23
The pathological process in bvFTD involves the frontal cortical structures and paralimbic areas, 24 -27 while the advanced stages involve further posterior temporal and parietal structures. 28 The structural changes in PPA are asymmetrical and primarily involve the left (dominant) hemisphere. In PNFA, the pathology is mainly localized to the convexity surface of the dominant insula and the dorsal frontal lobe, 29,30 while in the SD, the pathological process affects the anterior temporal lobe and limbic structures including the amygdala. 31,32 As the disease progresses the neurodegeneration of SD, spreads to other areas of both temporal lobes and insula, as well as to the ventromedial frontal areas, 33 while in PNFA, the frontal, superior temporal, and anterior parietal lobes are mainly affected. 34 The neurodegeneration of PSP is less clear. Progressive supranuclear palsy was considered to be associated mainly with degeneration of the deep brain structures—basal nuclei, brainstem, diencephalon, cerebellum and subcortical areas. 35 However, there is an increasing number of studies documenting that patients with PSP having cognitive impairment have cortical atrophy mainly involving frontal and temporal lobes. 36 -38
Despite this knowledge, the research concerning olfactory functions in FTLD subtypes is poor with inconsistent results. As in other neurodegenerative disorders, previous studies were focused mainly on odor identification. Although the odor identification deficit has been repeatedly shown in bvFTD 39 -43 and SD, 39,40,43,44 data concerning other FTLD subgroups are so far very contradictory. 40,41,43,45 -50 Because some FTLD subtypes are rare, all results are obtained from small sample sizes, and no study compares all FTLD subgroups. Similarly unclear is the nature of the odor identification deficit. Some authors hypothesized that there is a contribution of cognitive problems including memory 39,42 or executive dysfunction, 39 while other authors attributed the odor identification deficit to the damage of cortical olfactory structures. 40,41,43 A recent study suggested that odor identification deficit may be caused by disconnection between olfactory and cognitive areas. 45
The goal of this study was to determine or verify the presence and degree of odor identification impairment in different FTLD subtypes (bvFTD, SD, PNFA, and PSP) and to indicate the contribution of cognitive impairment to this deficit. Based on the supposed neuroanatomical substrate of each FTLD subtype and localization of olfactory pathways, we hypothesized that the odor identification may be impaired mostly in patients with bvFTD and SD and, to a lesser extent, in patients with PNFA and PSP.
Methods
Patients
Patients were recruited at the Memory Clinic, Department of Neurology, Charles University, 2nd Faculty of Medicine, Prague, Czech Republic, and the Memory Clinic, 1st Department of Neurology, St Anne's University Hospital, Brno, Czech Republic. All patients were nonsmokers and were screened to exclude conditions affecting olfactory functions. Patients with neuropsychiatric disorders, history of head trauma or maxillofacial surgery, with pathologies of the nose and paranasal sinuses, and alcohol or drug abuse were not included in the study.
The diagnoses were made as a consensus between neurologists and neuropsychologists and were based on clinical history, neurological and neuropsychological examination, and brain magnetic resonance imaging (MRI). Laboratory screening tests were performed to exclude secondary causes of symptoms. Three neurologists specialized in cognitive and movement disorders made a diagnosis based on a clinical history, neurological evaluation, MRI, and neuropsychological findings. In the case that all 3 experts reached the consensus, the diagnosis was accepted. In the case that one expert disagreed with others, the particular case was reanalyzed in detail and discussed again with neuropsychologists until consensus among all experts was reached.
Patients with a history of stroke, extensive vascular lesions on brain MRI (Fazekas visual scale with a score 2 or more), or a score of 7 or higher on the Hachinski Ischemic scale were not included in the study.
All patients had a history of gradually progressive symptoms. Diagnoses of bvFTD, PNFA, and SD were made based upon Neary's criteria. 13 Patients with bvFTD (n = 9) were characterized with dominant personality, behavior, and emotionality changes and decline in social interpersonal conduct. Patients with PNFA (n = 7) were characterized with prominent impairment of speech fluency, primarily including agrammatism, phonemic paraphasias, and anomia. Patients with SD (n = 6) were characterized with loss of word meaning and impaired recognition of objects and faces. Atrophy of frontal or/and temporal lobes on MRI was supportive for the diagnosis.
Patients with PSP (n = 8) were characterized mainly with postural instability with falls and vertical supranuclear palsy, and they fulfilled the
Patients included in the control group (n = 15) were age-matched healthy volunteers. All of them were nonsmokers. Exclusion criteria included a history of central nervous system diseases, psychiatric disorders, nose and paranasal sinuses pathologies, head trauma or maxillofacial surgery, alcohol or drug abuse, or other diseases or conditions potentially affecting olfaction. All patients gave a signed informed consent. The design of the study was approved by the institutional ethical committees.
Odor Identification Assessment
Odor identification was assessed using the Motol Hospital smell test (MHST)—a multiple-choice odor identification test 51 developed at Motol Hospital Memory Clinic. The MHST is composed of 18 flavors very well known among Czech older adults (pine-tree, peach, lemon, rose, cherry, grapefruit, clove, lavender, peppermint, orange, cinnamon, vanilla, coffee, honey, lilac, strawberry, black currant, and rum). Flavors were presented as essential oils, 200 μL each, in special phials to both nostrils simultaneously. After sniffing the odors, patients were asked to select a correct response from a 4-choice list, with 1 point given for each correct answer. The essential oils were replaced every 2 months in the phials in order to prevent degradation of the flavor.
In our previous study 51 involving 138 patients with FTLD, AD, and mild cognitive impairment and control participants, we documented that MHST results strongly correlate with the University of Pennsylvania smell identification test (UPSIT; r = .68, P < .0005). 52
In patients with SD and PNFA and in all patients suspected of having problems with comprehension, the picture identification test was administered to detect the nonolfactory elements of odor identification failure. Picture identification test represents a 40-item test analogous to the UPSIT where pictures instead of odors are used as stimuli. 53 No significant impairment was detected in any patient who was tested.
Neuropsychological Assessment
In all patients, general cognitive functions were evaluated by the Mini-Mental State Examination (MMSE). In all bvFTD, PNFA, SD, and control participants, and in 4 patients with PSP, neuropsychological testing was performed using a battery consisting of the Auditory Verbal Learning Test, the Free and Cued Selective Reminding Test, Trail Making Tests A and B, Digit Span forward and backward, the Controlled Oral Word Association Test, the Rey-Osterrieth Complex Figure, and the Boston Naming Test. Further assessments included the clinical dementia rating scale, activities of daily living (the functional activities questionnaire), and the geriatric depression scale. In the 4 other patients with PSP, who were recruited in the St Anne's University Hospital, a modified neuropsychological battery was performed including the frontal assessment battery, the semantic fluency test (animals, vegetables, colors, and cities), the Boston naming test, the free and cued selective reminding test, and the Montgomery-Asberg depression rating scale. All neuropsychological tests were adapted for the Czech population. All patients were Czech native speakers.
Statistical Analysis
Descriptive statistics involved a 1-way analysis of variance (ANOVA) to evaluate between-group differences in age, MMSE, and neuropsychological tests. The chi-square (χ 2 ) test was used to evaluate differences in proportions (gender). The between-group differences in the main odor identification analyses were evaluated using the 1-way ANOVA with post hoc analyses carried out with the Tukey’s honestly significant difference (HSD) test. A Pearson’s correlation was used to assess the relationship between the MHST and the neuropsychological tests. The data were screened for outliers, homogeneity of variance, and to ascertain whether the data were normally distributed. All data were found to be adequate for parametric analysis. In addition, we performed the nonparametric counterparts of parametric tests: Kruskal-Wallis test with subsequent Mann-Whitney U test with Holm-Bonferroni correction (counterpart of the 1-way ANOVA with the post hoc Tukey’s HSD test) and Spearman correlation (counterpart of parametric Pearson’s correlation) with the same results. For this reason, we reported only results from the parametric tests.
Further, we used multinomial logistic regression to assess whether the MHST could reliably distinguish patients with FTLD from the cognitively normal participants. We carried out the following specific steps: we defined the MHST scores to be independent variables, and then we entered the groups, that is, FTLD and controls, as 2 levels of the outcome, with the control group serving as the reference category. Results were expressed as odds ratios (ORs) and 95% confidence intervals (CIs), which corresponded to a 2-tailed .05 level of significance. The significance level in the descriptive statistics and in the main odor identification analyses was set at 2-tailed .05. All analyses were run using SPSS 13.0 for Windows.
Results
The groups were similar in age and gender but differed in global cognitive functioning measured by the MMSE (F 43,41 = 4.95, P = .003) and neuropsychological tests (Table 1). The post hoc analyses revealed that the control group outperformed the bvFTD (P = .048), PNFA (P = .009), SD (P = .024), and PSP (P = .034) groups in the MMSE and that the FTLD subgroups did not differ in MMSE. The group performances in neuropsychological tests are described in Table 1.
Characteristics of Study Participants.
Abbreviations: ANOVA, analysis of variance; bvFTD, behavioral variant frontotemporal dementia; FTLD, frontotemporal lobar degeneration; FCSRT, free and cued selective reminding test; HSD, honestly significant difference; PNFA, primary nonfluent aphasia; PSP, progressive supranuclear palsy; SD, semantic dementia; SD, standard deviation.
aAll scores are the raw scores.
bThe FTLD subgroups did not differ in neuropsychological tests with the exception of the FCSRT, where bvFTD and PSP groups outperformed the SD group (P < .05).
c P < .05 compared to the control group. The differences were evaluated using the 1-way ANOVA with post hoc Tukey’s HSD test.
d P < .01 compared to the control group. The differences were evaluated using the 1-way ANOVA with post hoc Tukey’s HSD test.
e P < .001 compared to the control group. The differences were evaluated using the 1-way ANOVA with post hoc Tukey’s HSD test.
The main analysis of odor identification showed the between-group differences in the MHST scores (F[4,41] = 5.69, P = .002). The control group outperformed the bvFTD (P = .015), SD (P = .035), PNFA (P = .017), and PSP (P = .026) groups (Figure 1). The differences between the FTLD subgroups were not significant.
Odor identification in frontotemporal lobar degeneration (FTLD) subtypes and control group.
Based on the finding that there were no differences among the FTLD subtypes in the MHST and to make the subsequent analyses more robust, we merged the FTLD subgroups and treated them as a 1 FTLD group.
Among the FTLD group, the MHST did not correlate with any neuropsychological measures. Results of the multinomial logistic regression analysis revealed that scoring 1 standard deviation below the mean on the MHST test was associated with almost 2 times greater odds of being classified as having FTLD compared to being cognitively normal (OR, 0.52; 95% CI, 0.34-0.80; P = .003).
Discussion
In this study, we investigated the presence and degree of odor identification impairment in FLTD subtypes (bvFTD, PNFA, SD, and PSP). The findings suggest that odor identification is impaired in all tested FTLD subgroups and that the severity of impairment does not differ across the FTLD subtypes. No correlation between odor identification impairment and cognitive domain-specific test was found.
Congruent with our results, previous studies reported that the odor identification deficit was present in patients with bvFTD 39 -43 and SD, 39,40,43,44 although some of the studies 43,44 were only case reports. Our results support and amplify these findings. This clear odor identification deficit in these subgroups may reflect the early involvement of structures that are crucial parts of the olfactory system—the frontal cortical areas in bvFTD and temporal limbic areas in SD.
On the contrary, the presence of odor identification impairment in the patients with PNFA remains unclear. Our results are consistent with 1 recent study 40 showing significant odor identification deficit in patients with PNFA and another reporting the same odor identification impairment in patients with PNFA and bvFTD. 41 On the other hand, 1 recent study 45 reported an isolated deficit of retrieving the names of odors in patients with PNFA (without problems with retrieving the names of objects or recognizing the name of the odor in a multiple-choice task); however, this study used only a short 8-item odor identification test, and the differences in odor identification scores (using the multiple-choice task) approached statistical significance. Similarly, a single case study 43 reported intact odor identification in a patient with PNFA compared to healthy participants, but the UPSIT score approached the cutoff for odor identification deficit (the patient scored 19 points of 40).
Also unclear is the extent to which structures that are affected early in PNFA are important for olfactory processing. Some studies 54 indicated that structures most vulnerable to damage in PNFA (insula and the dorsal frontal lobe) are essential for olfactory processing, but other studies 45 speculated that only associative pathways mainly located in the temporal pole are impaired. It may be that significant odor identification deficits in PNFA result primarily from more advanced structural changes.
Similar to patients with PNFA, the patients with PSP experienced odor identification deficits at the same level as the other FTLD subgroups. Previous studies of olfactory functions in patients with PSP yielded conflicting results. Some of the studies indicated intact or only mildly reduced olfactory functions, 46 -48 but others reported that olfactory functions are strongly reduced in patients with PSP. 49,50 These inconsistent findings may be the result of uneven impairment of cortical structures in patients with PSP. There is increasing evidence 36 -38 that patients with PSP presenting cognitive deficits have more pronounced frontal and temporal cortical gray matter loss, in contrast to cognitively intact patients with PSP. This finding is consistent with earlier studies reporting preserved olfactory functions in patients with PSP without cognitive impairment. 46,48 Further, a recent study found a strong association between odor identification impairment and cognitive deficits in patients with PSP. 50 Similarly, in our study, patients with PSP presented cognitive deficits (mean MMSE in our PSP group was 23.12). These data suggest that patients with cognitively impaired PSP may have the same degree of odor identification deficit as other FTLD subtypes.
We investigated odor identification without assessing other olfactory functions; and therefore, we cannot fully exclude a perceptual basis of odor identification failure in this cohort of FTLD patients. Previous studies reporting impairment of odor identification in the behavioral and language variants of FTLD and examining other olfactory functions (all of the studies were testing odor discrimination) 39,43,45 showed an isolated deficit of odor identification. There was no agreement concerning whether it was caused by cognitive failure or changes in olfactory structures. In patients with PSP, only a study 49 reporting 1 patient with PSP and evaluating all olfactory functions showed a significant olfactory deficit in this patient, but because the final score was a composite score from all olfactory subtests, it was not possible to determine the nature of the deficit.
Our study has several limitations. As with previous works dealing with olfaction in FTLD, our data come from a relatively small sample of patients, and our study is not supported with its own morphological data. Further neuroanatomical studies and confirmation with larger sample sizes should be carried out to support this hypothesis. We also cannot rule out that odor identification deficit in our patients with FTLD could be caused by impairment of odor detection or odor discrimination. Although there was no significant difference in gender among the groups, there were a higher number of women in the bvFTD, PNFA, and control groups and a higher number of men in the SD group. This might to some extent affect the results (might decrease the scores of SD and PSP groups when compared to the control group) as women are known to outperform men in smell identification ability. 55
In conclusion, our findings indicate that odor identification impairment is a feature common to different FTLD subtypes, and the presence of odor identification deficit may support the diagnosis of FTLD. The presence of olfactory impairment across all FTLD subgroups, together with no correlation between odor identification and neuropsychological findings, suggests changes in olfactory structures.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the European Regional Development Fund—Project FNUSA-ICRC (grant number CZ.1.05/1.1.00/02.0123); Grant Agency of the Czech Republic (grant number 309/09/1053); Internal Grant Agency of the Ministry of Health of the Czech Republic (grant number NT 11225-4); Institutional Support of Laboratory Research Grant No. 2/2012 (699002); Ministry of Health, Czech Republic—conceptual development of research organization, University Hospital Motol, Prague, Czech Republic 00064203; and Grant Agency of Charles University (grant numbers 74308 and 98509).